The Intelligent Selection Model Of Honing Parameters Based On Artificial Neural Network

The honing process is one of the high-precise finishing methods. The selection of the honing parameters has a great effect on the processing accuracy. Therefore how to choose the honing process parameters correctly is one of the issues remained discussing and researching by the honing technician. The traditional way of choosing honing parameters is to refer to the corrective manufacturing handbook or follow the working experience. As the technology and the application of the honing process develop, the variety of the process parts are more and more and the precision of the process parts are higher and higher. while the honing technical develops such as a honing tool,a honing stone and the honing process,the correct selection of the honing process parameters becomes more and more important. Aiming at this issue, with making use of the merits of the artificial neural network's fast convergence and high precision while solve the multiple objectives, analyzing the effect to the honing process quality by the material of the honing stone and process part and improving the algorithm of artificial neural network and optimizing the coding of both input and output, this research establishes the intelligent selection model of the honing parameters based on the artificial neural network. Then through the experimental research, this paper further validates the sophistication, feasibility and practicability of this intelligent selection model. Simultaneously, it lays the academic and experimental foundation for this method's practical application. This paper's chief contents as follows:Chapter 1. Generality: This chapter introduces orderly the origin of the research program, the research status at home and abroad and the chief contents of this research program. Through the investigation of the honing process statue at home and abroad, this research gives some new ideas of the development of the honing technology, and lays the foundation for the main research.Chapter 2. Analysis of the honing parameters: Through analyzing the typical honing parameters, investigating the honing specialist's experiences and collecting the documents about the honing process, this chapter preliminarily establishes the relationship between the process material, the process accuracy and honing parameters. This makes the arrangement for the construction of the artificial neural network model and the experiment for the sample data.Chapter 3. Forecast mode for honing parameters: Through the request of the selection of the honing parameters, this paper should plan to establish the mode and do the simulation respectively with several neural networks to contribute in reducing the variety and value range of the parameters. Thus this chapter designs the VB interface to join the multiple neural network modes and do some VB programming to simplify the mode structure. Based on this VB interface, the paper erects the neural network model and preliminarily fixes the parameters in the neural network model.Chapter 4. Honing experiment and neural network's simulation: According to the built structure of the neural network model, the sample data used for network's training is gained through the honing experiment combined with the specialist's experience. Then through the investigation on the sample data and research on the principle of the BP neural network, the neural network's structure is improved and the parameters in it are optimized. Finally by means of the measurement and simulation, it is certified that the intelligent selection model owns the sophistication on the technology and practicability on the production. It can reduce the experiment times evidently and normally is much better than the specialist's experience.Chapter 5.Summary: This chapter reviews and summaries the entire work, indicates the contribution, the innovation and the existing problems. Finally it prospects the further research.